NSF Awards: 2112532
The National AI Institute for Adult Learning and Online Education (AI-ALOE for short) is developing an AI-based transformative model for online adult learning. This model blends higher and continuing education with workforce capacity building to realize the potential of next-generation methods of instruction and assessment to support lifelong learning. These innovative transformations are not “just doing things better” but “doing better things” in effectiveness, efficiency, access, scale, and personalization. AL-ALOE is leading the country and the world in the development of novel AI theories and techniques for enhancing the quality of adult online education. It is fostering a research community of computer scientists to conduct responsible use-inspired fundamental research into AI that is grounded in theories of human cognition and learning, supported by evidence from large-scale data, evaluated on a large variety of testbeds, and derived from the scientific process of learning engineering. Together with partners in the higher education and educational technology sector, AI-ALOE is advancing online learning using virtual assistants to make education more available, affordable, achievable, and ultimately more equitable.
NSF Awards: 2112532
The National AI Institute for Adult Learning and Online Education (AI-ALOE for short) is developing an AI-based transformative model for online adult learning. This model blends higher and continuing education with workforce capacity building to realize the potential of next-generation methods of instruction and assessment to support lifelong learning. These innovative transformations are not “just doing things better” but “doing better things” in effectiveness, efficiency, access, scale, and personalization. AL-ALOE is leading the country and the world in the development of novel AI theories and techniques for enhancing the quality of adult online education. It is fostering a research community of computer scientists to conduct responsible use-inspired fundamental research into AI that is grounded in theories of human cognition and learning, supported by evidence from large-scale data, evaluated on a large variety of testbeds, and derived from the scientific process of learning engineering. Together with partners in the higher education and educational technology sector, AI-ALOE is advancing online learning using virtual assistants to make education more available, affordable, achievable, and ultimately more equitable.
Continue the discussion of this presentation on the Multiplex. Go to Multiplex
Chris Dede
Wirth Professor in Learning Technologies
We are excited to receive your advice and feedback on our work, which is in its early stages. We would also appreciate links to other initiatives you see as related, so we can make connections and learn from those resources.
Karen Royer
I am intrigued by the data sets that you describe in the video. You mention adult learning behaviors in the video. How do you plan on collecting your data sets? I have found that data on why adults engage in informal learning opportunities to be quite sparse. I think your work has the potential to make a great impact in this situation. What kinds of learning environments do you intend to include in your work?
Chris Dede
Wirth Professor in Learning Technologies
Karen, we are examining several approaches to collecting/mining/combining datasets. Caliper is one of the things we are studying; another is the new IEEE standards championed by ADL. We are interested in informal learning, but our early work is focused on formal instruction
Rachel Dickler
Hi Chris, thank you for sharing your video! It is exciting to see the work taking place across the AI institutes! There are really interesting intersections across our work at iSAT and AI-ALOE including the design of your virtual assistants (and ways in which they support instructors to create more equitable learning environments).
Shiran Dudy
Hi Chris, I am Shiran, a co-presenter of the Institute of Student-AI Teaming (iSAT), thanks for sharing your video and sharing the story of your institute! In many ways, I think we face similar challenges. I reach out since we are organizing a workshop where we gather together people who work in such collaborative environments, and in national NSF institute in particular, and are willing to share the challenges they face as part of working in an interdisciplinary team. We would love to hear from lessons learned of simply having you share your experience there. This is the workshop's website: Interdisciplinary Approaches to Getting AI Experts and Education Stakeholders Talking
Chris Dede
Wirth Professor in Learning Technologies
Shiran, thank you for sharing about the workshop. I will look at the website and consider how I could participate.
Shiran Dudy
Nathan Holbert
Associate Professor
Thanks for sharing team ALOE! I'm most curious about the ways in which your team and the network of AI researchers that you are cultivating on this project are exploring the online learning dataset. I know you all are in the early stages of this project, but are there there any particular insights or interesting patterns you've found in this data that can share? Or if you're not yet to that stage, can you say a bit about how you're interrogating this data?
And for other machine learning experts visiting this video, what questions or hypotheses would you be interested in exploring with such a data set?
Chris Dede
Wirth Professor in Learning Technologies
Nathan, thanks for asking We are still experimenting with different ways of structuring and interrogating online datasets. Our questions and hypotheses at this point center around developing AI-based instructional agents to aid human teachers.
Judi Fusco
Hi Chris,
Thanks for sharing the overview of your work with AI-ALOE. I look forward to learning more about what you're doing and hopefully have opportunities to connect around the work that AI Engage is doing. If there are writings you'd like to share with educators through the CIRCLS Educator blog, please let me know.
Good to "see" you and hear what you're doing!
Pati Ruiz
Chris Dede
Wirth Professor in Learning Technologies
Judi, I find CIRCLS really helpful. Will think about what I can share at this stage of our work.
Pati Ruiz
Joshua Danish
Professor and Program Chair
Hi Chris! Thanks for sharing and participating. Like so many of the folks in here, I am focused on younger learners. I wonder if you could say more about how your team views adult learners as similar to or different from younger learners? How will that shape the way you build on findings from other AI work, or how we might build on yours? (I'm part of the Engage AI institute).
Chris Dede
Wirth Professor in Learning Technologies
Joshua, thanks for asking. In our Institute, we are fortunate to have Professor Ruth Kanfer as our androgogy expert. There are stages within what we consider to be "adults" (age 24 and up), just as "children" have various developmental levels. I and Ruth would be happy to talk with you about what parts of our work can generalize to younger learners.
Joshua Danish
Professor and Program Chair
Thanks Chris! I was mostly trying to generate conversation but I'll definitely look forward to that once you have more data and lessons learned to share!
Maia Punksungka
Interesting work, Chris! I am curious to know more about the feedback you have received from adult learners who have used this model. How effective and user friendly has this program been to adult learners, specifically those who come from marginalized-minority communities or who are under educated? Are there any pre-requisites or basic skills that adult learners must possess prior to using this model?
Chris Dede
Wirth Professor in Learning Technologies
Maia, thanks for asking. We are in early implementations of our work, so don't yet have an answer for your question. However, we are focusing our initial development on learners who have the characteristics you describe: students at the Technical College System of Georgia. These studies will help us understand what will be most helpful to design.
Zenon Borys
Chris, what a great overview of a multifaceted project. As others have shared, there is so much potential. I find myself wondering about the datasets and their structure. You mentioned still experimenting with their structure. Could you share any surprising insights or challenges from figuring out how to structure them. I'm asking from a place where I see a dialectic relationship between what they tell use and how we have to format data going in. Knowing the structure shapes the data we collect/how what we have matters while what we find out makes new things matter (I think). Thanks!
Josh Sheldon
Project Lead
Hi Chris & all (lots of familiar faces in this discussion!),
This is a very interesting and ambitious project that prompts quite a few possible discussions.
I'm curious what questions/hypotheses others would pose for large datasets that are introduced in the video. Chris, can you provide any more detail about what categories of information those datasets will contain?
As well, what privacy and/or safety agreements and concerns might there be while assembling, storing, and making use of such datasets?
On the virtual teaching assistant front, I know you will be thinking hard about how to find the right balance between things humans can do uniquely or better than virtual assistants (for now) vs. the pieces an assistant can do which free up human instructors to do the things they do best.
But how do we determine what falls in which category, and what's more, how do we design systems that feed information from the assistant to human instructors to improve their interactions with learners? What do others think?
Such great opportunities for study!
Keep it up!
Chris Dede
Wirth Professor in Learning Technologies
Many of your questions center on how we are handling large datasets, which is a crucial component of our work. I am not an expert on this part of the Institute. For more detail, please contact our dataset expert: Professor Harshvardhan D Sikka <hsikka3@gatech.edu>. Thanks for your interest!